Tests on real datasets - Githubissues

polisquad / mpi-project

Implementation of the K-means algorithm with OpenMP and MPI

GNU General Public License v3.0

0 stars 1 forks source link

Tests on real datasets #8

Open sneppy opened 5 years ago

sneppy commented 5 years ago

Iris

$ time bin/main --init furthest --clusters 3 --convergence-factor 0.0625 --max-iter 256 data/iris.csv

real    0m0.008s
user    0m0.004s
sys     0m0.004s

iris

sneppy commented 5 years ago

R15

$ time bin/main --init furthest --clusters 15 --convergence-factor 0.0625 --max-iter 256 data/r15.csv

real    0m0.013s
user    0m0.048s
sys     0m0.004s

r15

sneppy commented 5 years ago

Aggregation

$ time bin/main --init furthest --clusters 7 --convergence-factor 0.0625 --max-iter 256 data/aggregation.csv

real    0m0.038s
user    0m0.188s
sys     0m0.000s

aggregation

sneppy commented 5 years ago

$ time bin/main --init furthest --clusters 15 --convergence-factor 0.0625 --max-iter 256 data/s1.csv

real    0m0.032s
user    0m0.112s
sys     0m0.008s

sneppy commented 5 years ago

$ time bin/main --init furthest --clusters 15 --convergence-factor 0.0625 --max-iter 256 data/s4.csv

real    0m0.028s
user    0m0.052s
sys     0m0.012s

sneppy commented 5 years ago

Unbalance

$ time bin/main --init furthest --clusters 8 --convergence-factor 0.0625 --max-iter 256 data/unbalance.csv

real    0m0.024s
user    0m0.032s
sys     0m0.000s

Bad run:

unbalance_bad

Good run:

unbalance_good